Back to Prompt Library
testing
Refining Audit Findings & Recommendations
Inspect the original prompt language first, then copy or adapt it once you know how it fits your workflow.
Linked challenge: AI Policy Audit Agent with OpenAI Agents
Format
Text-first
Lines
2
Sections
2
Linked challenge
AI Policy Audit Agent with OpenAI Agents
Prompt source
Original prompt text with formatting preserved for inspection.
2 lines
2 sections
No variables
0 checklist items
After the initial audit, instruct the agent to refine its findings and generate specific, actionable recommendations based on the identified violations. The agent should retrieve its previously saved findings from Mem0 to ensure continuity and leverage `GPT-4o`'s reasoning capabilities for detailed suggestions. Consider how OpenRouter could be configured to dynamically select a more cost-effective or specialized model for generating recommendations if applicable. For evaluation, provide a document that has a clear GDPR violation (e.g., lack of data minimization) and assess the agent's final output using the `RecommendationGeneration` task.
Adaptation plan
Keep the source stable, then change the prompt in a predictable order so the next run is easier to evaluate.
Keep stable
Preserve the rubric, target behavior, and pass-fail criteria as the baseline for evaluation.
Tune next
Adjust fixtures, mocks, and thresholds to the system under test instead of weakening the assertions.
Verify after
Make sure the prompt catches regressions instead of just mirroring the happy-path examples.